INTRODUCTION — Polycystic ovary syndrome (PCOS) is the most common cause of infertility in women [1], frequently becomes manifest during adolescence, and is primarily characterized by ovulatory dysfunction and androgen excess (hyperandrogenism). The syndrome is heterogeneous clinically and biochemically. The diagnosis of PCOS has lifelong implications with increased risk for metabolic syndrome, type 2 diabetes mellitus, and possibly cardiovascular disease and endometrial carcinoma. PCOS should be considered in any adolescent girl presenting with a chief complaint of hirsutism, treatment-resistant acne, menstrual irregularity, acanthosis nigricans, and/or obesity.

The diagnostic evaluation of an adolescent with suspected PCOS is described here. Other aspects of PCOS in adolescents are reviewed separately:

●(See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents".)

●(See "Treatment of polycystic ovary syndrome in adolescents".)

●(See "Etiology and pathophysiology of polycystic ovary syndrome in adolescents".)

INDICATIONS FOR EVALUATION — Work-up for PCOS is recommended for adolescent girls with one or more of the following characteristics:

●An abnormal degree of hirsutism or a hirsutism equivalent, such as inflammatory acne vulgaris, that is poorly responsive to topical therapies or oral antibiotics

●Focal hirsutism (localized areas of excessive sexual hair growth), if this is accompanied by menstrual abnormality

●Menstrual abnormality (persistent amenorrhea or oligomenorrhea, or excessive uterine bleeding) (table 1)

Acanthosis nigricans and/or obesity are manifestations of insulin resistance and metabolic syndrome and may be the presenting complaint of PCOS. Examination will usually reveal some of the above classical symptoms or signs, but occasionally these may not develop for some time. Thus, in the absence of risk factors for PCOS, such as a family history of PCOS, one may choose to defer a work-up of these latter patients for PCOS and instead observe them for the emergence of clinical evidence of PCOS over time. (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Clinical features'.)

EVALUATION OVERVIEW — We suggest a stepwise approach to the evaluation that addresses the diagnostic criteria (table 2) [2]. This approach is first outlined here and then detailed in the sections below (algorithm 1).

●Basic evaluation – The evaluation begins with a focused clinical evaluation for symptoms and signs suggestive of PCOS: menstrual irregularity, hirsutism (or hirsutism equivalents, such as acne), and acanthosis nigricans. This is followed by laboratory testing for hyperandrogenemia, starting with measurement of total or free testosterone as outlined in the algorithm (algorithm 1). Elevated total or free testosterone levels are very suggestive of PCOS. (See 'History and physical examination' below and 'Screening tests to detect common causes of abnormal menses' below.)

Patients with elevated testosterone should be further evaluated with a screening panel of laboratory tests to rule out other common non-PCOS causes of hyperandrogenemia (algorithm 2) [3]. This evaluation excludes the vast majority of disorders that mimic PCOS.

I also generally include ultrasonography of the ovaries and adrenal glands in the work-up of hyperandrogenic females, primarily to evaluate for rare tumors of these organs. Some other experts reserve the ultrasonographic screening for those patients with atypical features suggestive of virilization, such as rapidly progressive hirsutism, clitoromegaly, or hirsutism or menstrual abnormality that fails to respond to therapy. (See 'Screening tests to exclude common non-PCOS causes of hyperandrogenemia' below.)

In my pediatric endocrinology practice, I typically order most of these screening tests during the first visit. I take this approach because I consider it to be both simple and an economical use of time for most patients. However, measurement of serum 17-hydroxyprogesterone (17OHP) and other steroids requires early-morning sampling to be highly discriminatory in detecting nonclassic congenital adrenal hyperplasia (NCCAH). Some other experts perform the tests serially or select among the tests based on clinical symptoms and signs that raise concerns for a particular disorder.

●Diagnosis – Demonstration of persistent hyperandrogenism in a patient with a persistently abnormal degree of menstrual irregularity fulfills the diagnostic criteria for PCOS, provided that the endocrine screening tests are negative for disorders that mimic PCOS. This minimalist approach is approximately 99 percent specific for PCOS in young women and meets international consensus criteria for the diagnosis of PCOS in adolescents [4-7]. (See 'Diagnosis' below.)

Lack of hyperandrogenemia effectively rules out the diagnosis of PCOS in adolescents. However, the ovarian hyperandrogenism of PCOS may not become demonstrable until a few years after menarche [8,9]. Thus, young patients with menstrual irregularity should be followed, and the diagnosis of PCOS should not be dismissed until their menses normalize and androgen levels remain persistently normal.

●Further endocrine evaluation – The basic evaluation described above does not exclude rare virilizing disorders (table 3) [10-12]. In the presence of concern about the possibility of a rare virilizing disorder, such as a tumor or rare congenital disorder, a more comprehensive evaluation is suggested, including dexamethasone suppression testing and cosyntropin testing, which determine the source of androgen (algorithm 3). These tests also help to exclude the possibility that a mild PCOS picture is due to obesity and would be expected to respond to simple weight-control measures alone. This approach is consistent with the hirsutism clinical practice guidelines from the Endocrine Society [12]. (See 'Further endocrine evaluation for rare disorders mimicking PCOS' below.)

●Additional evaluation after the diagnosis of PCOS – Girls diagnosed with PCOS should have additional evaluations for glucose intolerance and other features of the metabolic syndrome, particularly if they are obese. Primary relatives of PCOS patients may also benefit from screening for these disorders. (See 'Additional evaluation of PCOS patients' below and 'Evaluation of family members' below.)

BASIC DIAGNOSTIC APPROACH

History and physical examination — The evaluation starts with a thorough assessment of the clinical symptoms and signs suggestive of hyperandrogenism and anovulation (ie, for disorders that mimic PCOS (algorithm 1)). The cutaneous manifestations of hyperandrogenism (particularly hirsutism and acne, which occur in approximately three-quarters of PCOS cases) provide clinical evidence of hyperandrogenism. An abnormal menstrual pattern constitutes evidence of oligo-anovulation.

●Hirsutism – The degree and distribution of sexual hair growth should be documented using the Ferriman-Gallwey score (figure 1). An adult level of hirsutism ordinarily is reached by two years after menarche [5]. Hirsutism should be interpreted in the context of norms for the patient's ethnicity. Moderate or severe hirsutism (hirsutism score >15) constitutes clinical evidence of hyperandrogenism in an adolescent [4-6]. The history should specifically explore whether the patient shaves excess hair or uses depilatory agents, which may obscure the physical findings, and whether the patient is taking medications that cause hirsutism (eg, anabolic-androgenic steroids, valproic acid). (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Hirsutism'.)

●Acne – The degree of acne should be evaluated in the context of the patient's gynecologic age. The possibility of hyperandrogenism is suggested by moderate or severe inflammatory acne (>10 lesions in any area, eg, face, chest, back (table 4)) through the perimenarcheal years or acne that is persistent and poorly responsive to topical or oral antibiotic dermatologic therapy [4-6]. (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Acne'.)

●Menses – The age of menarche and subsequent menstrual patterns should be described and interpreted in the context of the patient's gynecologic age. The types of menstrual dysfunction that suggest abnormal degrees of anovulation in adolescence are detailed in the table (table 1) [2]. A menstrual pattern that is outside of these bounds for two years (or one year with supporting evidence for PCOS) can be considered a "persistent" abnormality and fulfills one of the two criteria for the diagnosis of PCOS [2,4-6]. Note that normal menstrual regularity does not necessarily mean that ovulatory function is normal. (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Anovulation'.)

●Obesity and acanthosis nigricans – Obesity or acanthosis nigricans are often the initial complaint, though not the only PCOS symptom elucidated upon review of systems and examination. (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents".)

●Other – Assessment should include a history of medications that may either mask the symptoms (eg, oral contraceptives and topical or systemic acne medications) or cause the symptoms (eg, anabolic-androgenic steroids, valproic acid for epilepsy). Assessment should also include evaluation for features that suggest a hyperandrogenic disorder other than PCOS, including virilization (eg, rapidly progressive hirsutism), galactorrhea, Cushingoid or acromegaloid changes, evidence of thyroid dysfunction, or a family history of hyperandrogenic disorders (algorithm 1). (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Differential diagnosis'.)

Testing for hyperandrogenemia — Testing for testosterone and/or free testosterone should be included in the initial evaluation of a patient with clinical symptoms and signs suggesting PCOS (algorithm 1). However, these tests are only reliable for females if they are performed in a specialty laboratory using a highly sensitive and specific assay. Combined estrogen-progestin oral contraceptives (COCs) previously started should be withdrawn for two to three months before screening for PCOS because they suppress testosterone production. Current or recent use of systemic glucocorticoids also may moderately suppress total and free testosterone. Testing methods and interpretation are discussed below.

If a reliable testosterone assay is not available, the presence of moderate or severe hirsutism is approximately 85 percent specific for hyperandrogenism and may be used as clinical evidence for hyperandrogenism [4-6]. Mild hirsutism is a less reliable criterion for hyperandrogenism because only approximately one-half of patients with mild hirsutism have hyperandrogenemia, three-quarters of whom have a coincident menstrual abnormality [12].

For patients with severe or rapidly progressive hyperandrogenism, measurement of dehydroepiandrosterone sulfate (DHEAS) should be included in the initial evaluation to screen for a virilizing adrenal tumor.

Patients who have clinical features consistent with PCOS but have an initial normal total testosterone level, should have repeat testing measuring an early-morning (8 AM) serum free testosterone level in a reliable specialty laboratory by dialysis, or calculated from total testosterone and sex hormone-binding globulin (SHBG).

Considerations for testosterone assays — Persistent serum testosterone elevation above adult norms in a specialty reference laboratory is the best single test for hyperandrogenemia in postmenarcheal females [4-6].

The choice between testing for the presence of hyperandrogenemia with assay of total or free testosterone depends on assay reliability and cost-effectiveness considerations [12]. If a reliable measurement of total testosterone is easily available, this can be used as an initial test but should be followed by measurement of free testosterone if the results are not consistent with the patient's clinical course. If a reliable method for measuring serum free (or bioavailable) testosterone is available to the practitioner, and cost is not a significant issue, this test is the preferred choice for initial testing.

There are many pitfalls in testosterone assays at the low levels found in women, and reliable testosterone assays are not available to many clinicians. For these and other reasons discussed below, assessment may be best deferred to a specialist.

●Total testosterone – Reliable total testosterone assays are critical to accurate assessment for biochemical evidence of hyperandrogenism. This is important to obtain specificity for testosterone versus other steroids that cross-react in assays at the relatively low levels present in women and children. Their availability is limited to major specialty laboratories. Liquid chromatography-mass spectrometry methods are becoming the preferred method [13]. Where this technology is not available, high-quality postchromatographic radioimmunoassays yield comparable results and are the method of choice [6].

The automated assays that are used to measure serum total testosterone in most laboratories are not suitable to accurately measure levels in females [14,15]. Results by the best available assays vary by an average of approximately 6 to 26 percent [11,16]. The interpretation of the laboratory results is further complicated by systematic differences between assays and excessively broad normal ranges derived from populations of apparently normal women with unrecognized androgen excess.

●Free testosterone – An elevation of serum (or plasma) free testosterone is the single most sensitive test to establish the presence of hyperandrogenemia [4,12,15,17,18].

•The serum free testosterone concentration is approximately 50 percent more sensitive for the detection of hyperandrogenemia than the total testosterone concentration. This is because PCOS is characterized by low levels of SHBG, which is the main determinant of the fraction of serum testosterone that is weakly bound to albumin and the fraction that is free from protein binding, and thus ultimately bioactive [19,20]. Consequently, the combination of an upper-normal total testosterone and a lower-normal SHBG yields a high free testosterone concentration. SHBG production by the liver is raised by estrogen and hyperthyroidism, and it is suppressed in the settings of hyperandrogenemia and/or hyperinsulinemia (seen with obesity-induced insulin resistance) [21]. Although the low SHBG in obese individuals has been attributed to hyperinsulinemia [22], evidence suggests that excess glucose and fructose intake themselves and inflammatory cytokines mediate the SHBG reduction in patients with obesity [23].

•The most accurate free testosterone determinations come from specialty laboratories using established, validated assays in well-characterized control women. This is because lack of uniform assay methodology and standards introduces systematic differences between assays.

•Direct assays of the serum free testosterone concentration are inaccurate and should be avoided [12].

•The only reliable methods report free testosterone that is calculated as the product of the total testosterone and a function of SHBG (free testosterone = total testosterone × percent free testosterone) [12,24]. The most accurate methods calculate percent free testosterone as determined by equilibrium dialysis or, alternatively, as calculated from the SHBG concentration [25]. An alternative reliable method is to calculate "bioavailable testosterone" by determining the percent of serum testosterone not precipitated with globulins by ammonium sulfate (the supernatant includes both the fraction free and that weakly bound to albumin, the latter being bioactive to the extent that it dissociates to free in a given tissue space dependent upon the local interstitial albumin concentration) [24].

Interpretation of testosterone levels — The normal upper limit for serum total testosterone in adult women is approximately 40 to 60 ng/dL (1.4 to 2.1 nmol/L) when using most validated assays [15,17,18]. The adult norm is the appropriate reference range for adolescents [4,5]. Most patients with PCOS have serum testosterone concentrations of 29 to 150 ng/dL (1 to 5.2 nmol/L). A total testosterone >150 ng/dL (5.1 nmol/L) increases the likelihood of a virilizing ovarian or adrenal neoplasm: this cut-off has approximately 90 percent sensitivity and 80 percent specificity for androgen-producing tumors in adults [26].

For practical purposes, an elevated serum total and/or free testosterone at any time of day provides evidence of hyperandrogenism in an anovulatory cycle. However, a normal level in the afternoon does not exclude hyperandrogenemia. This is partly because serum testosterone undergoes episodic changes of approximately twofold (trough-to-peak). It is also partly because norms are standardized for early morning (8 AM) on days 4 through 10 of the menstrual cycle in regularly cycling women as normal testosterone levels fall 10 percent from 8 AM to 4 PM and double during midcycle [27,28].

The estrogen and progestin contents of COCs interfere with the assessment of androgens. They suppress gonadotropins, elevate SHBG, and directly inhibit steroidogenic enzymes such as 3-beta-hydroxysteroid dehydrogenase. They normalize androgens in PCOS and also have been reported to normalize androgens in some virilizing tumors [29]. After discontinuing COCs, normal women may transiently have a slightly high total testosterone level but a normal free testosterone level because SHBG turnover is slower than testosterone turnover.

Screening tests to detect common causes of abnormal menses — Most patients presenting for evaluation of hyperandrogenism have a menstrual abnormality for which screening for other causes of anovulatory disorders is indicated (algorithm 1). (See "Causes of primary amenorrhea" and "Epidemiology and causes of secondary amenorrhea".)

Initial screening for patients with a menstrual abnormality should include:

●Beta-human chorionic gonadotropin (beta-hCG) – Pregnancy should be excluded in all anovulatory patients.

●Chronic disease panel – Complete blood count, erythrocyte sedimentation rate or C-reactive protein (CRP), and comprehensive metabolic profile screen for chronic diseases that can cause anovulation.

●Follicle-stimulating hormone (FSH) and luteinizing hormone (LH) – Slightly elevated LH and slightly low FSH are characteristic of typical PCOS. Markedly elevated FSH suggests primary hypogonadism, whereas low LH indicates secondary hypogonadism (eg, a pituitary tumor). Glucocorticoids suppress gonadotropin levels.

●Prolactin – Hyperprolactinemia is a cause of gonadotropin deficiency; it is not accompanied by galactorrhea if the gonadotropin deficiency is profound.

●Thyroid-stimulating hormone (TSH) – Normal serum TSH levels are ordinarily adequate to rule out thyroid dysfunction, which causes menstrual dysfunction. Hypothyroidism may cause confusion with hyperandrogenism by causing coarsening of hair (which can be mistaken for hirsutism [5]), lowering SHBG, and causing multicystic ovaries. Hyperthyroidism may confound the diagnosis of hyperandrogenism by raising SHBG, which may elevate the total testosterone. Note that mildly elevated levels of TSH are often found in obese individuals, but this is a consequence rather than a cause of the obesity and is reversible if weight loss can be achieved. (See "Acquired hypothyroidism in childhood and adolescence".)

Screening tests to exclude common non-PCOS causes of hyperandrogenemia — If serum testosterone and/or free testosterone is elevated, the next step is to exclude most non-PCOS causes of hyperandrogenemia. Adult and adolescent specialty society guidelines for the diagnosis of PCOS differ slightly as to which tests should be routinely obtained versus which tests should be considered only in those patients with suggestive clinical features of disorders that may mimic PCOS [4,30-33]. Consequently, expert practice varies regarding the extent and timing of further laboratory evaluation. In my practice, I routinely perform a simple screening battery of endocrine tests and ultrasonography at this stage (algorithm 2 and table 3). Other experts perform these tests in selected patients with atypical features, such as virilization (to rule out virilizing neoplasm), central obesity (to rule out Cushing syndrome), or acromegaloid features.

Endocrine screening panel — In patients with PCOS, the following endocrine studies will be normal. An abnormal result, however, for any of these tests suggests another cause of hyperandrogenism (see "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Differential diagnosis') and should be further evaluated as suggested in the linked topic reviews.

●Early-morning 17-hydroxyprogesterone (17OHP) – An 8 AM 17OHP level is a good screening test for nonclassic congenital adrenal hyperplasia (NCCAH) secondary to 21-hydroxylase deficiency when performed under properly controlled circumstances:

•Early-morning sampling is critical to detect the 17OHP elevation of NCCAH because it wanes rapidly thereafter due to the diurnal variation of adrenal steroid secretion.

•It is important to obtain the screening sample when the patient is amenorrheic or within the first 10 days after the start of a menstrual cycle in regularly cycling patients because 17OHP rises during the preovulatory and luteal phases of the cycle. A 17OHP value of >170 ng/dL (5.1 nmol/L) is suggestive of NCCAH in an anovulatory cycle but is also compatible with recent ovulation. This cutoff displayed approximately 95 percent sensitivity and 90 percent specificity in detecting NCCAH, with the main confounder being PCOS (>20 percent of patients with PCOS have elevated baseline 17OHP concentrations, as do some with tumoral hyperandrogenism) [34-37]. An abnormal screening test is not diagnostic but requires a cosyntropin (ACTH) stimulation test to confirm the diagnosis of CAH [38], unless the basal 17OHP level is >1000 ng/dL (30 nmol/L) [38]. (See 'Further endocrine evaluation for rare disorders mimicking PCOS' below.)

•I also measure serum progesterone at the same time as the 17OHP measurement to rule out the possibility that the patient has unexpectedly ovulated (which occurs in approximately 10 percent of oligo-amenorrheic PCOS patients) and is unknowingly being tested in the luteal phase of her cycle. The luteal phase is indicated by a serum progesterone >175 ng/dL (5.5 nmol/L) and can be accompanied by 17OHP levels up to 350 ng/dL (10.5 nmol/L).

•In those patients with known CAH that is complicated by PCOS (see "Etiology and pathophysiology of polycystic ovary syndrome in adolescents", section on 'Congenital virilization'), full glucocorticoid replacement treatment will suppress the 17OHP level to normal, but androgen levels will remain slightly elevated. (See "Etiology and pathophysiology of polycystic ovary syndrome in adolescents", section on 'Congenital virilization'.)

●Other androgens – DHEAS is included primarily to screen for an adrenal tumor. Routine testing for other androgenic steroids, such as androstenedione, has been of little diagnostic utility in most populations [12]. It is a marker for adrenal hyperandrogenism and has little diurnal variation. While the most common cause of DHEAS elevation is the functional adrenal hyperandrogenism of PCOS [39], the main purpose of measuring DHEAS levels is to rapidly identify an unusual virilizing adrenal disorder, such as cortisone reductase deficiency [40,41] or an adrenal tumor. Girls with a virilizing tumor usually present with a rapid onset of virilizing features, and DHEAS levels are often, but not necessarily, markedly elevated (>700 mcg/dL, 19 micromol/L) if the tumor is of adrenal origin. CAH seldom causes severe DHEAS elevation [39]. However, in a substantial minority of patients with virilizing tumors, the symptoms are indolent in onset and mimic PCOS in presentation (see "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Differential diagnosis' and "Adrenal hyperandrogenism"). COCs have equivocal effects on the DHEAS level. (See "Physiology and clinical manifestations of normal adrenarche".)

The measurement of other androgens, such as androstenedione, is ordinarily of little diagnostic utility, except in situations of diagnostic uncertainty [12].

●Prolactin – Prolactin should be measured at this stage if it was not done in the initial panel of screening tests for abnormal menses. Androgen excess occurs in 40 percent of hyperprolactinemic women due to multiple effects of prolactin excess on adrenal androgen production and androgen metabolism, and the great majority (approximately 85 percent) of these women have galactorrhea [42]. The combination of hirsutism, galactorrhea, and amenorrhea is sometimes described as Forbes-Albright syndrome. Prolactin elevation is unusual in PCOS itself, occurring in less than 1 percent of subjects, and marginal elevation in the absence of specific clinical evidence does not necessitate a prolactinoma work-up [33,43]. Serum prolactin values more than 25 ng/mL usually have an identifiable cause rather than PCOS [43]. Estrogenic medications increase prolactin secretion, but the amount of estrogen in low-dose COCs generally does not cause hyperprolactinemia. (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Differential diagnosis' and "Clinical manifestations and evaluation of hyperprolactinemia".)

●Serum cortisol – Serum cortisol is an optional test that many experts perform only for hyperandrogenic patients with central obesity. The purpose is to screen for endogenous Cushing syndrome (due to adrenal hyperplasia or adrenal tumors), which is sometimes associated with adrenal overproduction of testosterone. Although the normal range for serum cortisol is wide (5 to 25 mcg/dL), a serum cortisol concentration of <10 mcg/dL (276 nmol/L) is reassuring evidence against endogenous Cushing syndrome. A midday or afternoon sample is more optimal for this screening than an early-morning sample. Patients with markedly elevated serum cortisol levels or those with clinical features suggestive of Cushing disease warrant further evaluation (eg, with 24-hour urine collection for free cortisol and creatinine). Estrogen raises total serum cortisol levels by increasing cortisol-binding globulin, but low-dose COCs do not generally cause falsely abnormal results. (See "Establishing the diagnosis of Cushing's syndrome".)

●Insulin-like growth factor 1 (IGF-1) – This is an optional test that many experts perform only for hyperandrogenic patients with gigantism or acromegaloid features. Elevated IGF-1 distinguishes growth hormone (GH) excess as a cause of hyperandrogenism [44] from the pseudoacromegalic hyperandrogenism of severe insulin resistance syndromes [45]. GH excess usually can be identified by clinical symptoms (gigantism in growing children or acromegaly after epiphyseal fusion) in teenagers and adults. However, the author has seen acromegaly first present with PCOS-like symptoms. The oral estrogen of COCs may lower IGF-1 levels sufficiently to obscure the diagnosis of acromegaly. (See "Diagnosis of acromegaly".)

Ultrasonography — The primary purpose of ultrasonography in the hyperandrogenemic adolescent is to exclude causes other than PCOS. A secondary benefit is to identify the few individuals with very large ovaries, for whom it simply provides further evidence of PCOS severity and diagnostic specificity, as discussed below. Ultrasonography is neither recommended nor required for the diagnosis of PCOS in adolescents, because the high frequency of polycystic-appearing ovaries in normal females in this age group makes this an unreliable criterion for the diagnosis of PCOS [46].

●The role of ultrasonography in the differential diagnosis of PCOS – Practice varies as to the indications for ultrasonography in girls with confirmed hyperandrogenemia since ultrasonography is not recommended for the diagnosis of PCOS. In my pediatric endocrinology practice, I have recommended ovarian and adrenal ultrasonography for all patients with anovulatory symptoms and documented hyperandrogenemia to exclude rare but serious androgen-producing tumors, as described below. Other experts perform ultrasonography only for selected patients with features that are atypical for PCOS, such as very high testosterone levels (eg, >150 ng/dL), clitoromegaly, rapidly progressive hirsutism, or poor response to treatment [47].

The primary purpose of ultrasonography is to screen for the rare but serious ovarian tumor (algorithm 2). In addition, other pelvic pathology, including an ovotesticular disorder of sex development and the functional hyperandrogenism of pregnancy, may be detected by ultrasonography. On rare occasions, ultrasonography has been insensitive in detecting a virilizing ovarian tumor in adults [48,49]. It may also reveal a virilizing adrenocortical tumor and, if there is reason to suspect this, more specialized imaging studies are indicated (see "Clinical presentation and evaluation of adrenocortical tumors"). Girls who have an ultrasound that shows an ovarian tumor or other explanation for hyperandrogenism should be referred for further evaluation and treatment of the underlying disorder. Otherwise, hyperandrogenic girls need further endocrine studies irrespective of whether the ovaries are polycystic, as discussed in the section above. (See 'Endocrine screening panel' above.)

Ultrasonography also provides the opportunity for patient reassurance and education. For many women, the diagnosis of ovarian "cysts" raises a concern about tumors, so it is reassuring to know that a tumor has been ruled out by the ultrasound. The clinician can explain to the PCOS patient with a polycystic ovary that these are numerous small egg sacs that are not ovulating properly. Conversely, if a polycystic ovary is not visualized in a PCOS patient, the clinician can explain that the ovarian dysfunction is "too mild to be seen on the ultrasound."

●Ultrasonographic criteria for polycystic ovary morphology (PCOM) – The ultrasonographic finding of PCOM is supportive of a diagnosis of PCOS, but it is not included in the 2015 diagnostic criteria for PCOS in adolescents [4-6]. This is because of uncertainties about the definitive criteria for PCOM in this age group. However, there is reason to believe that the presence of clear PCOM, in combination with symptomatic hyperandrogenism in the absence of anovulatory symptoms, poses a risk for PCOS in an adolescent [50]. Therefore, we suggest careful follow-up of girls with PCOM. In adults, these same features would meet a somewhat nonspecific PCOS criterion. (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Adults'.)

International consensus criteria for adults define PCOM on the basis of either excessive size or follicle number (or both) in the absence of a dominant-sized follicle (>1 mL) or a corpus luteum, as evaluated by transvaginal ultrasonography (image 1) [32,51] (see "Diagnosis of polycystic ovary syndrome in adults", section on 'Transvaginal ultrasound'). On the one hand, there is a high prevalence of PCOM in the most clinically severe PCOS cases [50]. On the other hand, PCOM is a common normal variant in asymptomatic women that may predict a slightly longer period of fertility [50]. Furthermore, it has become apparent that these criteria are problematic in young adults because they naturally have slightly larger ovaries [52], and the newer high-definition vaginal imaging techniques show that small antral follicle counts up to 24 are normal [53].

Adult PCOM criteria are especially problematic when applied to adolescents [4,5]. For one thing, an antral follicle count cannot be accurately defined by the abdominal ultrasonographic approach necessary in virginal adolescents [53,54]. Additionally, ultrasound studies suggest that ovarian volume is slightly larger in adolescents than in adults, though data vary considerably [4,50,54]. Consequently, one-quarter to one-half of normal adolescents meet Rotterdam adult criteria for PCOM [54,55]. In 2015, an international consensus group proposed using a (mean) ovarian volume >12 mL to constitute PCOM in adolescence [4]. However, this threshold may be too low since a single ovarian volume up to 14 mL was the median of three ultrasonographic studies of normal adolescents [50].

These ultrasound data are supported by two magnetic resonance imaging (MRI) studies of well-characterized, healthy, normally menstruating adolescents and adolescents with PCOS, totaling approximately 40 per group [56,57]. They reported that the upper normal cutoff for small antral follicle counts per ovary is up to 17 to 25 and concurred that the upper normal for the size of a single ovary is 14 mL.

DIAGNOSIS — PCOS can be diagnosed in a patient with a persistently abnormal uterine bleeding pattern and evidence of hyperandrogenism (table 2) after exclusion of other causes of most non-PCOS causes of hyperandrogenemia (table 3). Most clinical guidelines recommend screening for nonclassic congenital adrenal hyperplasia (NCCAH), Cushing's syndrome, prolactin excess, thyroid dysfunction, and acromegaly. However, the guidelines vary as to which tests are recommended for universal screening and which are recommended only for patients with symptoms suggestive of one of these disorders. These guidelines are based on expert opinion. Therefore, the actual range of practice varies considerably among expert clinicians in this field.

It is also best to confirm the persistence of hyperandrogenism in order to avoid labeling as PCOS adolescents who have transient hyperandrogenemia due to physiologic adolescent anovulation [4] (see "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Differential diagnosis'). Approximately one-quarter of adolescents with an abnormal degree of menstrual irregularity will have an elevated androgen level but no clinical manifestations of hyperandrogenism, and neither the hyperandrogenemia nor the menstrual abnormality will persist. In the absence of clinical signs of hyperandrogenism (eg, hirsutism or an abnormal degree of acne), a diagnosis of PCOS is not warranted in an adolescent with anovulatory symptoms and an elevated androgen level without other clinical manifestations of PCOS, unless both of these features are documented to persist for two years [4,5].

Adolescents with PCOS or with features of PCOS that do not fulfill diagnostic criteria should be followed to determine that hyperandrogenemia and symptoms persist since data are limited on the natural history of adolescent PCOS. The diagnosis of PCOS should not be fully dismissed until menses normalize and androgen levels remain persistently normal.

FURTHER ENDOCRINE EVALUATION FOR RARE DISORDERS MIMICKING PCOS — Some endocrinologists perform additional testing in hyperandrogenemic patients to detect disorders that mimic PCOS and would go undetected by the above basic screening approach. These disorders include the approximately 1 percent or less who have a rare virilizing disorder that may be encountered no more than a few times in an endocrinologist's career (table 3), and the 8 percent in whom simple obesity may account for a mild PCOS picture in the absence of ovarian androgenic dysfunction [11]. This diagnostic approach is consistent with that recommended by hirsutism guidelines from the Endocrine Society for evaluation of hirsute women with hyperandrogenemia [12], but the cost effectiveness is unknown. (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Differential diagnosis'.)

Our suggested approach to a more comprehensive diagnostic endocrine evaluation (algorithm 3) is typically performed only in selected patients during an evaluation by a subspecialist (eg, a pediatric or reproductive endocrinologist). This evaluation may be particularly helpful in patients with atypical features, such as virilization (eg, rapidly progressive hirsutism), unexplained congenital or familial hyperandrogenism, or unresponsiveness to standard therapy. Our approach augments the initial evaluation by starting with a dexamethasone androgen-suppression test (DAST) to distinguish an adrenocorticotropic hormone (ACTH)-dependent adrenal source of androgen excess from other sources of androgen (algorithm 3) [10]. If indicated by these tests or atypical clinical or laboratory findings, further work-up may also include specialized imaging studies such as computed adrenal tomography [12]. (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Other' and "Clinical presentation and evaluation of adrenocortical tumors".)

Dexamethasone androgen-suppression test — The responses of androgens (testosterone and dehydroepiandrosterone sulfate [DHEAS]) and glucocorticoids to dexamethasone suppression are the primary outcome measures to establish the underlying source of hyperandrogenism (algorithm 3). Normal responses in our laboratory are given in the algorithm legend.

●Baseline levels of steroid intermediates, such as 17-hydroxypregnenolone (17OHP), 11-deoxycortisol, and androstenedione, are very elevated in rare forms of congenital adrenal hyperplasia (CAH) and some virilizing tumors. Baseline 24-hour urine glucocorticoids (ie, free cortisol and 17-alpha-hydroxycorticosteroids) with creatinine are helpful for the unusual case of Cushing's syndrome and the rare case of cortisone reductase deficiency. The latter is characterized by elevated urinary 17-alpha-hydroxycorticosteroid excretion that is comprised predominantly of cortisone rather than cortisol metabolites [40].

●A short (four-hour) DAST can be used to screen for an ovarian source of androgen in obese women [11]. This consists of a noontime 0.5 mg oral dose of dexamethasone followed by blood sampling four hours later for cortisol, DHEAS, and testosterone. A short DAST is 95 percent sensitive for detecting the functional ovarian hyperandrogenism of PCOS. Therefore, normal suppression of plasma testosterone by a short DAST rules out PCOS with high probability. However, subnormal testosterone suppression in response to a short DAST cannot be expected to distinguish virilizing disorders or Cushing's syndrome from PCOS.

●A long (four-day) DAST is the definitive form of dexamethasone suppression testing for the differential diagnosis of hyperandrogenic disorders.

●The DAST is interpreted as follows:

•If testosterone excess is not suppressed by dexamethasone, but cortisol and DHEAS are suppressed normally, the diagnosis of PCOS is virtually assured. However, virilizing tumor and adrenal rests must still be considered in the presence of suggestive clinical factors.

•If neither androgen (testosterone and DHEAS) excess nor glucocorticoids are suppressed normally by dexamethasone, then disorders other than PCOS must be considered, such as Cushing's syndrome, adrenal tumors, and glucocorticoid resistance (or dexamethasone was not taken properly). If the levels of androgens suggest an adrenal tumor, further imaging, such as computed tomography, is indicated.

•If both androgens (testosterone and DHEAS) and glucocorticoids are suppressed normally, an ovarian source for androgen is unlikely, so further evaluation for adrenal disorders is indicated with a cosyntropin (ACTH) test for CAH.

Cosyntropin (ACTH) stimulation test — The responses of steroid intermediates on the pathway to cortisol and androgens distinguish adrenal steroidogenic disorders that mimic PCOS (algorithm 3).

●The cosyntropin (ACTH) stimulation test (using a 250 mcg dose, with samples drawn 30 to 60 minutes later) is interpreted as follows:

•A 17OHP value >1000 ng/dL (30 nmol/L) is highly suggestive, and >1500 ng/dL is definitive for the 21-hydroxylase deficiency form of nonclassic congenital adrenal hyperplasia (NCCAH) [58]. Intermediate 17OHP levels (1000 to 1500 ng/dL) require genetic confirmation to make a definitive diagnosis of NCCAH [37,59]. (See "Diagnosis and treatment of nonclassic (late-onset) congenital adrenal hyperplasia due to 21-hydroxylase deficiency".)

•Mildly elevated responses of 17-hydroxypregnenolone and DHEA are characteristic of the primary functional adrenal hyperandrogenism of PCOS and are often confused with nonclassic 3-beta-hydroxysteroid dehydrogenase deficiency (3-beta-HSD) [60]. Nonclassic 3-beta-HSD is a consideration only if the 17-hydroxypregnenolone response to ACTH is >10 standard deviation (SD) above the normal mean (ie, >4500 ng/dL) [150 nmol/L]) [61]. (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Differential diagnosis' and "Etiology and pathophysiology of polycystic ovary syndrome in adolescents", section on 'Functional adrenal hyperandrogenism'.)

•Nonclassic 11-beta-hydroxylase deficiency is a rare mime of PCOS [62]. Only an 11-deoxycortisol response to ACTH >5 times greater than the upper limit of normal (>4000 ng/dL/116 nmol/L) has been shown to be specific for mutation detection. (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Differential diagnosis'.)

If both the DAST and the ACTH stimulation test are normal, obesity or idiopathic hyperandrogenism is most likely. The most common cause of this may be the pseudo-PCOS of obesity [11] (see "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Differential diagnosis'). The rare possibility of cortisone reductase deficiency should be considered on the basis of the clinical picture. (See "Dexamethasone suppression tests" and "Initial testing for adrenal insufficiency: Basal cortisol and the ACTH stimulation test".)

Ancillary tests

●Serum levels of anti-müllerian hormone (AMH) are mildly increased in subjects with polycystic ovary morphology (PCOM; a category that includes both normal adolescents and PCOS subjects) [55,63,64], while AMH elevations of twofold or more above the upper normal limit are highly specific for PCOS [63]. AMH is a product of the granulosa cells of small growing follicles that normally restrains follicular growth and sex steroid secretion [65,66]; thus, it acts as an intraovarian paracrine gatekeeper to control the recruitment of resting primordial follicles into the growth phase [55,63,64]. Androgen excess in the ovary stimulates primordial follicle growth, causing an increase in antral follicle count that contributes to PCOM. Thus, AMH levels are independently related on the one hand to the size of the follicle pool ("follicle reserve" [63,67]) and PCOM (for which AMH has been suggested as a surrogate for ultrasonographic criteria to define PCOM), and on the other hand hyperandrogenism. (See "Etiology and pathophysiology of polycystic ovary syndrome in adolescents", section on 'Primary functional ovarian hyperandrogenism'.)

●Gonadotropin-releasing hormone (GnRH) agonist test – In patients with PCOS, administration of a GnRH agonist induces a hyperresponsive elevation of 17OHP without evidence of a steroidogenic block. The steroid pattern in response to the test distinguishes PCOS from other causes, particularly rare hyperandrogenic disorders of steroidogenesis, such as nonclassic 3-beta-hydroxysteroid dehydrogenase deficiency. The test is performed by administering leuprolide acetate 10 mcg/kg subcutaneously and measuring steroid levels 20 to 24 hours later; the test can be performed in conjunction with dexamethasone administration (as part of the DAST described above) to suppress coincidental adrenal secretion that might interfere in the interpretation [68,69]. This test may facilitate the early diagnosis of PCOS in adolescents [70]. It also helps in determining if mild PCOS is simply due to obesity, which is suggested when hyperandrogenemia is mild, and serum luteinizing hormone (LH), AMH, and DHEAS, along with ovarian morphology, are normal [11]. (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Differential diagnosis' and "Uncommon congenital adrenal hyperplasias", section on '3-beta-hydroxysteroid dehydrogenase type 2 deficiency'.)

ADDITIONAL EVALUATION OF PCOS PATIENTS — Once a diagnosis of PCOS has been established, the possibility of insulin-resistance manifestations should be considered. This is because PCOS is a risk factor for the early development of type 2 diabetes mellitus, metabolic syndrome, and their associated risks, such as sleep-disordered breathing [71,72]. Approximately one-quarter of teenagers with PCOS meet proposed adolescent criteria for the metabolic syndrome. Patients should also be evaluated for quality-of-life issues. (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Manifestations of insulin resistance' and "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Psychological issues'.)

We advise screening for type 2 diabetes mellitus in adolescents with PCOS and obesity, or other risk factors for diabetes mellitus. At the least, a fasting plasma glucose or hemoglobin A1c should be measured. However, an oral glucose tolerance test (OGTT) is preferable because it is the most sensitive and specific measure of glucose tolerance (see "Epidemiology, presentation, and diagnosis of type 2 diabetes mellitus in children and adolescents", section on 'Laboratory tests'). This recommendation is consistent with those made in an international reproductive endocrinology consensus publication [73], the American Association of Clinical Endocrinologists (AACE), and the minority position of the Androgen Excess and PCOS Society (AE-PCOS) in adult women with PCOS [74,75]. The prevalence of diabetes was reported to be 2 percent based on the fasting blood sugar in one series, and 8 percent when based on OGTT criteria in another series of adolescents [76,77]. In both series, almost all of the adolescents had no symptoms of diabetes. Impaired glucose tolerance suggests insulin resistance and is a risk factor for type 2 diabetes mellitus and cardiovascular disease. Thus, an abnormal OGTT has important therapeutic implications. (See "Treatment of polycystic ovary syndrome in adolescents", section on 'Obesity and insulin resistance'.)

Glucose tolerance should be monitored regularly because a substantial number of women with PCOS will experience deterioration in glucose tolerance. As an example, among 25 adolescent and young women followed for a mean of 34 months, the two-hour plasma glucose during an OGTT increased at an average rate of 9 mg/dL (0.5 mmol/L) per year [78]. Among the 14 women with PCOS and normal glucose tolerance at baseline, 55 percent experienced deterioration of glucose tolerance when they were retested with an OGTT. Among the 14 women with PCOS and impaired glucose tolerance at baseline, 29 percent progressed to diabetes.

Patients should also be evaluated for quality-of-life issues. (See "Definition, clinical features, and differential diagnosis of polycystic ovary syndrome in adolescents", section on 'Psychological issues'.)

PCOS is a risk factor for endometrial carcinoma in young women [79]. The basis of the risk is multifactorial: it is related to the endometrial hyperplasia that arises from persistent estrogen stimulation without the progesterone-induced inhibition of proliferation and differentiation to secretory endometrium that occurs after ovulation. Hyperandrogenism, insulin-resistant hyperinsulinemia, and the inflammatory changes of obesity appear to be aggravating factors for endometrial carcinoma in women with PCOS; risk also seems related to body mass index (BMI)-independent proto-oncogenic changes in endometrial cells [80]. (See "Clinical manifestations of polycystic ovary syndrome in adults", section on 'Endometrial cancer risk'.)

EVALUATION OF FAMILY MEMBERS — Parents and siblings of PCOS patients are at increased risk for metabolic syndrome and diabetes mellitus, particularly if they are obese. Screening can be accomplished by measurement of hemoglobin A1c or oral glucose tolerance testing (OGTT) in first-degree relatives of either sex. Premenopausal mothers and sisters are at risk for PCOS and should be screened for those features.

These recommendations are prompted by the high prevalence of PCOS and metabolic syndrome among immediate relatives of individuals with PCOS (see "Etiology and pathophysiology of polycystic ovary syndrome in adolescents", section on 'Heritable traits'). According to one study, approximately one-half of sisters of PCOS probands have an elevated serum testosterone level, and one-half of these (one-quarter of the total) in turn have menstrual irregularity and thus meet National Institutes of Health (NIH) criteria for PCOS [81]. In another study, approximately one-quarter of sisters met the Androgen Excess and PCOS Society (AE-PCOS) criteria for PCOS, having hyperandrogenism and a polycystic ovary, although menses were ovulatory [82]. The likelihood of a mother having PCOS seems similar: 22 percent in our series [76].

The prevalence of metabolic syndrome or diabetes mellitus in parents of women with PCOS is higher than in parents of women without PCOS. In our series, one-third or more of mothers and the great majority of fathers had metabolic syndrome and diabetes mellitus [76]. Notably, the diabetes was asymptomatic in one-half and was uncovered by glucose tolerance testing. Meta-analysis has demonstrated a significantly greater prevalence of type 2 diabetes mellitus and insulin resistance in the parents of PCOS patients and a trend to diabetes and insulin resistance in siblings [83].

OTHER RESOURCES — The following online resources are available to patients with PCOS and their families:

●PCOS resources for a healthier you – From the Center for Young Women's Health of Boston Children's Hospital [84]

●Polycystic Ovary Syndrome: A Guide for Families – From the Pediatric Endocrine Society (PES) and the American Academy of Pediatrics (AAP) [85]

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Polycystic ovary syndrome" and "Society guideline links: Hirsutism".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topic (see "Patient education: Polycystic ovary syndrome (The Basics)")

●Beyond the Basics topic (see "Patient education: Polycystic ovary syndrome (PCOS) (Beyond the Basics)")

SUMMARY

●The evaluation for polycystic ovary syndrome (PCOS) begins with a focused clinical evaluation for the presence of hirsutism (or hirsutism equivalents, such as acne that is resistant to topical or oral antibiotic therapy) and menstrual abnormality (table 1). Some patients may present with a chief complaint of acanthosis nigricans or obesity before the more classical manifestations develop. (See 'Indications for evaluation' above.)

●The clinical evaluation is followed by laboratory testing for common causes of abnormal menses, including testing for androgen excess with total or free testosterone (algorithm 1). (See 'History and physical examination' above and 'Screening tests to detect common causes of abnormal menses' above.)

●If serum free testosterone levels are normal (in the absence of oral contraceptives), the diagnosis of PCOS is unlikely. However, the possibility of PCOS in an adolescent should not be fully dismissed until menses normalize and androgen levels are persistently normal. (See 'Testing for hyperandrogenemia' above.)

●If serum levels of total and free testosterone are elevated, a focused history and physical examination should be performed to exclude other hyperandrogenic disorders. Expert practice varies regarding the extent and timing of further laboratory evaluation. In my practice, I perform a simple screening battery of endocrine tests and ultrasonography at this stage (algorithm 2 and table 3). Other experts measure 17-hydroxyprogesterone (at 8 AM) in many or all patients but perform the other screening tests only in patients with atypical features, such as Cushingoid features, galactorrhea, or virilization. (See 'Screening tests to exclude common non-PCOS causes of hyperandrogenemia' above.)

•If the results of the endocrine screening panel are normal, the diagnosis of PCOS is confirmed with a high level of certainty (table 2). Any abnormal results suggest that the hyperandrogenism is caused by a disorder other than PCOS. (See 'Endocrine screening panel' above and 'Diagnosis' above.)

•The primary purpose of ultrasonography is to exclude the rare but serious virilizing ovarian tumor that can mimic PCOS. Determining whether the ovaries are polycystic is not helpful for the diagnosis of PCOS, because of the high frequency of polycystic-appearing ovaries in adolescents with or without PCOS. (See 'Ultrasonography' above.)

●Patients with atypical features, such as virilization, unexplained congenital or familial hyperandrogenism, or unresponsiveness to standard PCOS therapy may have a rare virilizing disorder rather than PCOS. In such cases, we suggest evaluation for rare disorders mimicking PCOS, including the endocrine screening panel described above (if not already done) and a dexamethasone androgen-suppression test (DAST) (algorithm 3). (See 'Further endocrine evaluation for rare disorders mimicking PCOS' above.)

●Once a diagnosis of PCOS has been established, it is important to identify and monitor for abnormal glucose tolerance, type 2 diabetes, and other features of the metabolic syndrome. This is because PCOS is a risk factor for the early development of these disorders. Immediate family members are also at risk for these metabolic dysfunctions, and premenopausal mothers and sisters are at risk for PCOS. (See 'Additional evaluation of PCOS patients' above and 'Evaluation of family members' above.)